One-way ANOVA

The ANOVA test is based on the cleaned data from regression step, dedicated to show the association between variables and the outcome.

Rows: 25116 Columns: 23
── Column specification ─────────────────────────────────────────────────
Delimiter: ","
chr  (11): month, boro, location_desc, perp_age_group, perp_sex, perp...
dbl  (10): incident_key, day, year, precinct, jurisdiction_code, x_co...
lgl   (1): statistical_murder_flag
time  (1): occur_time

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

showing = function(x){
  x %>% 
  broom::tidy() %>%
  mutate(
    p.value = format(p.value, scientific = TRUE, digits = 3)
  ) %>% 
  select(term, p.value) %>% 
  rows_delete(tibble(term = "Residuals")) %>% 
  rename(variable = term) %>% 
  knitr::kable()
}

Year

res_year = aov(number_shoot ~ factor(year), data = nyc_fit_lm)
res_year_df = showing(res_year)

Matching, by = "term"

The p-value of the year variable is smaller than 0.05, thus we reject the null hypothesis and regard the year as statistically significant.

Month

res_month = aov(number_shoot ~ factor(month), data = nyc_fit_lm)
showing(res_month)

The p-value of the month variable is smaller than 0.05, thus we reject the null hypothesis and regard the month as statistically significant.

Borough

nyc_boro = nyc_fit_lm %>% 
  rename(borough = boro)
res_boro = aov(number_shoot ~ factor(borough), data = nyc_boro)
showing(res_boro)

The p-value of the borough variable is smaller than 0.05, thus we reject the null hypothesis and regard the borough as statistically significant.

Statistical_murder_flag

res_smf = aov(number_shoot ~ factor(statistical_murder_flag), data = nyc_fit_lm)
showing(res_smf)

The p-value of the Statistical_murder_flag variable is smaller than 0.05, thus we reject the null hypothesis and regard the Statistical_murder_flag as statistically significant.

Perp_sex

res_perpsex = aov(number_shoot ~ factor(perp_sex), data = nyc_fit_lm)
showing(res_perpsex)

The p-value of the sex of perpetrator variable is smaller than 0.05, thus we reject the null hypothesis and regard the sex of perpetrator as statistically significant.

Perp_race

res_perprace = aov(number_shoot ~ factor(perp_race), data = nyc_fit_lm)
showing(res_perprace)

The p-value of the race of perpetrator variable is smaller than 0.05, thus we reject the null hypothesis and regard the race of perpetrator as statistically significant.

Vic_sex

res_vicsex = aov(number_shoot ~ factor(vic_sex), data = nyc_fit_lm)
showing(res_vicsex)

The p-value of the sex of victims variable is smaller than 0.05, thus we reject the null hypothesis and regard the sex of victims as statistically significant.

Vic_race

res_vicrace = aov(number_shoot ~ factor(vic_race), data = nyc_fit_lm)
showing(res_vicrace)

The p-value of the race of victims variable is smaller than 0.05, thus we reject the null hypothesis and regard the race of victims as statistically significant.

Perp_age_group

res_perp_age_group = aov(number_shoot ~ factor(perp_age_group), data = nyc_fit_lm)
showing(res_perp_age_group)

The p-value of the perp_age_group variable is smaller than 0.05, thus we reject the null hypothesis and regard the perp_age_group as statistically significant.

Vic_age_group

res_vic_age_group = aov(number_shoot ~ factor(vic_age_group), data = nyc_fit_lm)
showing(res_vic_age_group)

The p-value of the vic_age_group variable is smaller than 0.05, thus we reject the null hypothesis and regard the vic_age_group as statistically significant.